Rstudio covariance structure analysis error persists (lavaan function)

What I'm trying to do
I would like to analyze the covariance structure using Rstudio's lavaan package (cross-delay effect model)
The data I'm using is panel data from two points in time

Current Code

model2009<-'

# Two potential variables, internal, installation, predict five observational variables: INFLU, COMPLEX, PARTY, ELECTION, CONGRESS (one or two after the variable name represent which of the two data points).
## Constraints coefficients to identify models
internal1=~1*INFLU1+a*COMPLEX1
installation1 = ~1 * PARTY1 + b * ELECTION1 + c * CONGRESS1
internal2=~1*INFLU2+a*COMPLEX2
installation2=~1*PARTY2+b*ELECTION2+c*CONGRESS2

# This is the error part of the measurement variable.
INFLU1~INFLU1; COMPLEX1~ COMPLEX1; PARTY1~ PARTY1~ PARTY1; ELECTION1~ELECTION1; CONGRESS1~CONGRESS1;
INFLU2~INFLU2; COMPLEX2~ COMPLEX2; PARTY2~ PARTY2~ PARTY2; ELECTION2~ELECTION2; CONGRESS2~CONGRESS2;

# This is the structural equation.
internal2 through internal1 + installation1;
installation2 through internal1 + installation1;

# This is the error part of the latent variable.
internal1~~installation1; internal2~~installation2; internal1~~internal1; installation1~~installation1; internal2~internal2~internal2~installation2~installation2
'

# We predict these models using the lavaan() function.
result2009<-lavaan(model2009, panel.2009, ordered=c("PARTY1", "ELECTION1", "CONGRESS1", "INFLU1", "COMPLEX1", "PARTY2", "ELECTION2", "CONGRESS2", "INFLU2", "COMPLEX2"),
                     WLS.V = TRUE, sample.nobs = 1107)

Problems you are facing
When I try to function with the above model, R generates the following error

Error in Gamma.g*tcrossprod(a1): Incorrect array

I tried searching, but I'm not sure what this error is.

Notes
I tried going back with the taraceback() function, but I didn't really understand the content

traceback()
4:lav_test_satorra_ventler_trace_ABA(Gamma=Gamma, Delta=Delta, 
       WLS.V = WLS.V, E.inv = E.inv, ngroups = ngroups, nobs = lavsamplestats@nobs, 
       total=lavsamplestats@ntotal, return.ugamma=return.ugamma, 
       Satterthwaite= Satterthwaite)
3:lav_test_satorra_ventler(lavobject=NULL,lavsamplestats=lavsamplestats, 
       lavmodel=lavmodel, lavimplied=lavimplied, lavdata=lavdata, 
       lavoptions=lavoptions, TEST.unscaled=TEST[[1]], E.inv=attr(VCOV, 
           "E.inv"), Delta=attr(VCOV, "Delta"), WLS.V=attr(VCOV, 
           "WLS.V"), Gamma=attr(VCOV, "Gamma"), test=this.test, 
       mimic=lavoptions$mimic, method="ABA", return.ugamma=FALSE)
2:lav_model_test(lavmodel=lavmodel,lavpartable=lavpartable, 
       lavpta=lavpta, lavsamplestats=lavsamplestats, lavimplied=lavimplied, 
       lavh1 = lavh1, lavoptions = lavoptions, x = x, VCOV = VCOV, 
       lavdata=lavdata, lavcache=lavcache, lavloglik=lavloglik)
1:lavaan(model2009, panel.2009, ordered=c("PARTY1", "ELECTION1", 
       "CONGRESS1", "INFLU1", "COMPLEX1", "PARTY2", "ELECTION2", 
       "CONGRESS2", "INFLU2", "COMPLEX2"), WLS.V = TRUE, sample.nobs = 1107)

2022-09-30 21:50

1 Answers

The cause was due to arguments in the lavaan function.

Details

The reason for the resolution is that you removed the WLS.V argument in the lavaan function (this time it is sem).
WLS.V should be TRUE if it is a correlation matrix instead of raw data, but at the time of the first question, it was TRUE even though it was using raw data, so a warning message or error occurred.

Resolved Code

model_cross2009<-'
# Since we use sem() instead of lavaan(), we do not need to write the error variance of measurement variables and potential variables.
internal1 = through 1*INFLU1 + COMPLEX1;
installation1 = ~ PARTY1 + ELECTION1 + CONGRESS1;
internal2=~1*INFLU2+COMPLEX2;
installation2 = ~ PARTY2 + ELECTION2 + CONGRESS2;

internal2 through internal1 + installation1;
installation2 through internal1 + installation1
'

result_cross2009<-sem(model_cross2009, 
                        panel.2009, ordered=c("Party 1", "ELECTION 1", 
                                               "CONGRESS 1", "INFLU1", 
                                               "COMPLEX 1", 
                                               "PARTY 2", "ELECTION 2", 
                                               "CONGRESS 2", "INFLU2", 
                                               "COMPLEX 2"

2022-09-30 21:50

If you have any answers or tips

Popular Tags

python x 4647

android x 1593

java x 1494

javascript x 1427

c x 927

c++ x 878

ruby-on-rails x 696

php x 692

python3 x 685

html x 656